BACKGROUND: In 2012, the US Food and Drug Administration (FDA) issued the draft guidance \"Determining the Extent of Safety Data Collection Needed in Late Stage Premarket and Postapproval Clinical Investigations.\" The selective data collection approach proposed in this guidance leads to reductions in costs and work time and may improve the quality of the database for clinical trials. The current study evaluated the applicability of selective data collection for oncology drugs.
METHODS: The labeling information of oncology drugs obtained supplemental approvals from the FDA between 2005 and 2014 were used. The frequency of adverse reactions observed in clinical trials between the first approval and supplemental approvals of a specific drug were compared. Paired studies were categorized into the following 4 groups: A, same tumor type and same usage; B, same tumor type and different usage; C, different tumor type and same usage; D, different tumor type and different usage.
RESULTS: A total of 46 study pairs for additional drug indications were investigated. In group A, 6 of the 7 pairs showed a high correlation coefficient ( r = 0.988, 0.953, 0.947, 0.935, 0.853, and 0.846).
CONCLUSIONS: Selective data collection should be adopted in cases in which the additional indication of a drug is for the same tumor type and usage as the first or previous indication.